metadata
dataset_info:
features:
- name: task
dtype: string
- name: neg
dtype: string
- name: query
dtype: string
- name: pos
dtype: string
- name: margin
dtype: float32
splits:
- name: train
num_bytes: 159353054.6377521
num_examples: 151649
download_size: 90263657
dataset_size: 159353054.6377521
configs:
- config_name: default
data_files:
- split: train
path: data/train-*